In alko989/s6model: Data-poor stock assessment using a single-sepcies steady-state size-based (s6) model

Introduction

The usage of s6model R-package is illustrated using the North Sea cod (Gadus morhua) as an example. The status of this data-rich stock is assessed using only fisheries independent information, i.e. size frequency distributions from scientific surveys. Data from the North Sea International Bottom Trawl Survey (NS-IBTS) are downloaded from the DATRAS database. Further, total yearly landings and the results from the official assessment are also obtained from ICES.

The North Sea cod (Gadus morhua) is assessed in this example using only survey data from the NS-IBTS survey. The data can be downloaded from the DATRAS database.

Downloading and pre-processing of the data

The data are downloaded using the DATRAS package which can be installed using:

install.packages('DATRAS',repos='http://www.rforge.net/',type='source')

The package contains functions to download, read and subset survey data. Shape files for ICES regions can be downloaded from ICES spatial facility.

library(DATRAS)
## Download the survey data for the years 1991 to 2014
downloadExchange("NS-IBTS", 1991:2014)
## Read in all data
nsibts <- readExchangeDir()
## Select only cod
nscod <- subset(nsibts, Species == "Gadus morhua")

Select only the relevant areas for the North Sea Baltic cod

## Add area information
nscod <- addSpatialData(nscod, shape = "ices_areas.shp")
nscod <- subset(nscod, ICES_area %in% c("IIIa", "IVa", "IVb", "IVc", "VIId"))

Then using the s6model package we can extract the yearly weight frequencies using the function getDFYears. This function extracts the length distributions from an object of class DATRASraw, fits the weight length relationship, transforms the length to weight and returns a list of data.frames, one for each year. The bin size is set to 100 g (binsize argument) and the weight-length relationship is fitted using weight and length data in the survey for each year (estWL argument).

library(s6model)
nscoddat <- getDfYears(nscod, binsize = 100, estWL = TRUE)

Now the data are in a format that can be used to assess the stock status. The data are provided with the package (saved as nscoddat).

Assessment of the stock

Now the data are in the correct format the assessment can be done using the makeAssessment function. It has many arguments and only a subset of them are presented here. There are default values for most of them, so that a preliminary assessment can be done using makeAssessment(dat), where dat has the data in the correct format. The results are saved after a successful run of the function so that they can accessed at a later point to make plots and reports.

Since it is an assessment using only survey data, we are using isSurvey = TRUE and we make the assumption that all sizes are selected by the gear by setting the relative selectivity parameter ($\eta_S$, argument eta_S) of the gear very small. The results are not saved (fnout = NULL). The default values for mean and sd of physiological mortality from @Kokkalis2016 are used.

library(s6model)
yrs <- s6model:::getYears(names(nscoddat))
w <- ices.cod[["cod-347d"]]$Year %in% yrs
land <- ices.cod[["cod-347d"]]$Landings[w]
res <- makeAssessment(nscoddat, fnout = NULL, isSurvey = TRUE, eta_S = 1e-3, yield = land, u = 10,
                      a = 0.22, a.sd = 0.7, nsample = 20, binsize = 10)

Reporting the results

The results of the assessment of $F/F_{msy}$ are plotted using plot function. The y-axis is in log-scale and its limits are set.

par(mar = c(3.1, 3.1, 0.1, 0.1))
s6model:::plot.s6modelResults(res, ylim = c(1/15,15), log = "y")
addIces("cod-347d", col = 1)
plot(res, what = "ssb", mult = 1/1000, ylim = c(0,130), addhline = FALSE)
addIces("cod-347d", mult = 1/1000, what = "SSB")

alko989/s6model documentation built on Nov. 2, 2023, 10:04 p.m.